…otherwise the classpath validator will complain the the scala library is missing, therefore it is working as expected :)
…matically" This reverts commit 18a034b.
… will cause the ContentAssistError dialog to pop up. I also changed usages of 'pos.start' into 'pos.startOrPoint' as it is safer. This fix only prevents the error dialog to show up, nd it does not fix ticket Re #1000705 (completion is still not working as expected).
…th is invalid. The reset in this case is useless, generates log and was making the tests fail on trully headless systems.
It looks like my fix for #1000668 was not enough; the refactoring library gets confused quickly if it doesn't have a completely typechecked AST. This also fixes #1000599. (cherry picked from commit 2296fd8)
…tion compiler was not yet initialized. This avoids waking up the compiler just to tell it to shutdown, and also eliminates a source of deadlocks or recursive initializations (for instance, if reset is called while initializing the presentation compiler). Fixed #1000704.
… builder bugs
…er instead. Fixed #1000357.
…c search works, but only through the Java Search window (no right-click support). Method definitions and references are not distinguished, probably still some bugs that need ironing out. In the medium term we need to revamp ScalaMatchLocator. Refs #1000698.
…r 2.8 compatibility
…s that usage of 'Manifest.singleType', which is available only for 2.9+. To fix compilation I duplicated in our codebase the implementation of 'Manifest.singleType` and added a deprecation note for removing the change once we abandon 2.8 support (maybe we would be better of with a custom assertion...). This change is related to ticket Re #1000691. Commit should be reviewed by @mdr." This reverts commit 2d82c80.
… few more test for checking that Java code can correctly reference trait's types.
…sing module's inner classes to JDT. The issue was that module's inner classes were exposed as member of the $ module class, which is wrong, since the scala compiler expose them as inner classes of the module's companion class. I moved the code that was responsible of traversing the Scala AST Tree into a new class 'TreeTraverser', that takes a new field 'module2innerClassDefs' used to keep track of inner classes that are defined in a module. This solution has the advantage of avoiding possible memory leaks that could result from not correctly cleaning the 'module2innerClassDefs' map. Fixed ticket Re #1000678.
…spite Martin seem to have fixed it in the Presentation Compiler. I think we should NOT close the ticket until we know why the test is failing (we'll take a look at it after 2.0 release).
I must have missed this when I went through all the refactoring code. This seems to fix #1000668, where the refactoring generated invalid code. (cherry picked from commit 81ae1c4)
…sage of 'Manifest.singleType', which is available only for 2.9+. To fix compilation I duplicated in our codebase the implementation of 'Manifest.singleType` and added a deprecation note for removing the change once we abandon 2.8 support (maybe we would be better of with a custom assertion...). This change is related to ticket Re #1000691. Commit should be reviewed by @mdr.
…use the issue needs to be fixed in the presentation compiler (I tried to reproduce the problem with a presentation compiler test, but so far I did not succeeded in isolating the issue outside the IDE. Though, I'm sure that the issue described in this ticket has to be fixed in the presentation compiler.). While creating the test I had to refactor a bit class 'HyperlinkDetectorTester'. Specifically, I needed to extract the functionality used to test hyperlinking in a scala file, as the issue described in this ticket becomes apparent only *after* a hyperlink request is issued on a specific class' symbol. Last, I also changed CustomAssertion to have a self type 'TestProjectSetup', as in this way the current 'project' does no longer need to be passed to 'assertNoErrors'.
…00664' into release/scala-ide-2.0.x